Interoperability Minutes 2007-02-07

From MemberWiki

Jump to: navigation, search

Full minutes: /member/wiki/Interoperability_Minutes_2007-02-07

Contents

OpenAjax Alliance Interoperability Committee meeting minutes 2007-02-07

Attendees

  • Adam Peller <apeller(at)us.ibm.com>
  • Alex Russell <alex(at)dojotoolkit.org>
  • Craig McClanahan <craig.mcclanahan(at)sun.com>
  • Gideon Lee <glee(at)openspot.com>
  • Ingo Muschenetz <ingo(at)aptana.com>
  • James Margaris <jmargaris(at)nexaweb.com>
  • Jesse Costello-Good <jcostell@tibco.com>
  • Jon Ferraiolo <jferrai(at)us.ibm.com>
  • Ken Tam <kentam(at)bea.com>
  • Kevin Hakman <khakman(at)tibco.com>
  • Phil Berkland <berkland(at)us.ibm.com>
  • Rama Gurram <rama.gurram(at)sap.com>
  • Rick Evans <Rick.Evans(at)Sun.COM>
  • Shailesh Hingole <shingole(at)gce2000.com>
  • Ted Thibodeau <tthibodeau(at)openlinksw.com>

Original Agenda

Minutes

Topic: Transition from Interoperability Committee to Interoperability Working Group

Jon: Transtioning to formal process, must follow reccomendation and form working group. Need working group before releasing approved spec. Starting work on charter for working group. Have draft chater, trying to narrow down to not invade patent space. Discuss charter in e-mail and at face-to-face. Review early, especially companies with IP concerns.

Kevin Hackman: Not going to have finished spec until april or may?

Jon: Correct, no approved spec until then. Must develop plan for going to 1.0. Something to discuss. My thinking get approved in summer.

Topic: Next F2F meeting - March 22-23

Jon: Meeting march 22-23 at IBM office. Have starting agenda and Wiki where you can add your name. Want to minimize presentation and status reports, maximize problem solving and work. Wiki has topics also, you can add your own topics of interest to it. Topics include OpenAjax1.0, review of interop-fest, talk about open source infrastructure. Expecting concrete proposals from IDE and server task force at this meeting?

Topic: Review proposed changes to Hub Spec, open source reference implementation and InteropFest test case

Jon: On to proposed changes from Adam Peller. Change undefined check to foo = foo || {}?

Craig McClanahan: We should comment some of the JS idioms, people not familiar with them. Separate out commented version and compressed version?

Alex: JS in code not that adventurous.

Jon: Probably need to distinguish between commented out and compressed stuff at some point?? (I missed where this ended up, just trailed off)

Jon: Removing argument checks?

Alex: Are we supporting polymorphic calls? Doesn't look like it, that is the main use case.

Jon: Put them in for casual dev, can be handy. But size is a major issue.

???: Difference between just using wrong arg types and passing null or bad length. Runtime vs write-time error. What about a debug js file?

Alex: Could throw debug stuff in separate file that wraps functions in closure.

Jon: Sounds good. What do people think about that?

James: Throwing errors can sometimes obfuscate the problem as well, in favor of that.

???: Priority on this?

Jon: Low probably. Someone can write the code if it is that important to them. Removing argument checks sounds final.

???: Protect access? Is that important?

??? (tibco?): Obfuscating code makes the difference here less big maybe.

???: What if we want registering stuff has side effect, fires event?

Jon: I experimented with something like that. notified on library adds.

James: If we aren't cloning the returned array they can mess with it even with a getter.

???: At some point we had a clone of the array.

Jon: We would have to say array is read only in spec. Do people read spec?

James: Most people's gut reaction is to not mess with array. Or at least mine.

Alex: I would remove without a second thought, we will be shipping this everywhere.

???: Public API more flexible.

Alex: Can always listen for array push pop etc.

Jon: Vote...sounds unanimous.

Jon: Collision check - we don't want to run this every time do we?

Alex: Similar to arg checking, split off into separate file?

Jon: Probably should be separate files, debug and collision check.

Craig: Makes sense to me.

Jon: Bug in load/unload.

Jon: Loosening up test cases around load/unload, some script needs to be loaded in body after onload() has been called.

Alex: Difference between injecting script tag after the fact vs. writing it in body itself.

??? (tibco): Set a timeout and write the scrip tag there.

Alex: Is there a variable you can check to see if document is stable?

Jon: We can set loaded variable and they can check it.

Kevin: onload event can get clobbered.

Jon: Is loading dynamically in the body acceptable? It should be.

Alex: Need to be resilient to window.onload changes.

???: Browsers already have a way to add load/unload handlers without clobbering. Why are we duplicating that?

Jon: Library logic, app logic. Need library/ components to run first.

Jon: Changing test case to handle dynamic loading sounds approved.

Jon: Remove load handlers after load has occured? Might as well. No need to keep them around.

Jon: What about 'hub has been loaded' global?

James: Doesn't track document iteself, just state of hub, whether or not it has initialized, not whether or not document itself is loaded.

Jon. Correct, should not name as 'documentLoaded'.

Jon: Let's continue talking about spec next time, will get hub changes done by next week hopefully.

Retrieved from "/member/wiki/Interoperability_Minutes_2007-02-07"